Month: February 2017

The \”middle-income trap\” is an argument that when countries have emerged from dire poverty to middle-income status, they can become stuck at that point, and stop making progress toward higher income levels. The World Development Report 2017  notes: \”Contrary to what many growth theories predict, there is no tendency for low- and middle-income countries to converge toward high-income countries.\” The overall theme of this year\’s WDR is  \”Governance and the Law,\” and as usual, the report offers a wealth of examples and insights. Here, I\’ll just focus on the arguments about the middle-income trap, where the report illustrates its underlying theme by arguing that \”the difficulty many middle-income countries have in sustaining growth can be explained by power imbalances that prevent the institutional transitions necessary for growth in productivity.\”

This figure illustrates the patterns of transition for economies between low-income, middle-income, and high-income status. On the horizontal axis, countries are plotted by their per capita income level in 1970; on the vertical axis, by their per capita income level in 2010.

To get a sense of how the graph works, look at the category of \”lower-middle income\” countries on the horizontal axis, with per capita GDP between 5 and 15% of the world leaders. Now run your eye up, and see how those countries are faring by 2010. A substantial share of them have fallen into the \”low-income\” category, although most remain in the same \”lower middle\” category as before. Only one of thee lower-middle countries from 1970, Korea, had emerged into the high-income category after 40 years. Similarly, if you start by looking at low-income countries in 1970, only two of them had risen as high as \”upper middle income\” by 2010: Equatorial Guinea (GNQ) and Botswana (BWA), which is a prosperity largely founded on oil and diamonds, respectively.

The World Bank researchers writing the WDR argue that a core problem is that the institutions and strategies that raise a country up to middle-income status are often different from the strategies that would allow taking the next step to high-income status–and entrenched interest groups can make the transition a difficult one. Here\’s some commentary from the WDR (citations and references to figuress omitted):

\”Middle-income countries may face particular challenges because growth strategies that were successful while they were poor no longer suit their circumstances. For example, the reallocation of labor from agriculture to industry is a key driver of growth in low-income economies. But as this process matures, the gains from reallocating surplus labor begin to evaporate, wages begin to rise, and decreasing marginal returns to investment set in, implying a need for a new source of growth. Middle-income countries that become “trapped” fail to sustain total factor productivity (TFP) growth. … Efficient resource allocation and industrial upgrading require a set of institutions that differs from those that enable growth through resource accumulation. … 

\”The creation of these institutions may be stymied by vested interests. Creative destruction and competition create losers—and in particular may create losers of currently powerful business and political elites.This is a more politically challenging problem than spurring productivity growth through the adoption of foreign technologies, which tends to favor economic incumbents. These political challenges may be particularly great in middle-income countries because actors that gained during the transition from low to middle income may now be powerful enough to block changes that threaten their position. In this sense, the challenges that middle-income countries face go beyond policy choice to the challenge of power imbalances. … Understanding the policy arena in which elites bargain is essential for explaining the political economy traps faced by middle-income countries.

\”One such political economy trap is a persistent deals-based relationship between government and business. Deals-based, sometimes corrupt, interactions between firms and the state may not prevent growth at low income levels; indeed, such ties may actually be the “glue” necessary to ensure commitment and coordination among state and business actors. But they become more problematic for upper-middle-income countries. For example, theory suggests that as markets expand and supply networks become more complex, deals-based relationships can no longer act as a substitute for impersonal, rules-based contract enforcement. … Combating entrenched corruption and creating a
level playing field for firms imply a need for accountable institutions. At upper-middle-income levels, legislative, judicial, media, and civil society checks become increasingly important.\” 

The difficulties in moving toward types of governance that can offer a foundation for both representation and growth is an ongoing theme throughout the report. As another example, here are some facts about elections worldwide that raised my eyebrows. The share of countries holding elections is steadily rising, but the the share of elections rates as \”free and fair\” is steadily falling.

The number of elections is steadily rising (as shown by the bars) but voter turnout worldwide in those elections has been steadily falling.

Governments have become much less likely to censor the media in a direct way in the age of the internet, but they have become more aggressive about regulations that limit the  ability of civil service organizations (CSOs) to organize themselves or to spread their messages.  The report notes:

As the report notes:

\”Evidence from the last decade, however, suggests that the global trend may be a shrinking civic space (figure 8.10). Many governments are changing the institutional environment in which citizens engage, establishing legal barriers to restrict the functioning of media and civic society organizations, and reducing their autonomy from the state. For example, in the case of media, governments may award broadcast frequencies on the basis of political motivations, withdraw financial support of media organizations and activities, or enforce complex registration requirements that raise barriers to entry into a government-controlled media market. In the case of nongovernmental organizations (NGOs), governments might resort to legal measures to restrict public and private financing or pass stricter laws that restrain associational rights …\”

In short, economic growth and development isn\’t just about pulling the right economic policy levers–government budgets, monetary policy, investment in education, foreign aid, and the like. It\’s also about the extent to which economic forces have flexibility to function within the political and legal institutions of that society.

For some earlier posts on the hurdles in the way of economic convergence, see

\”Will Convergence Occur?\” (November 25, 2015)
\”Dani Rodrik on Economic Convergence: Jackson Hole I\” (September 14, 2011)

A typical intro-level conception of the economy points to the household sector as net savers, who then through the financial system end up financing the investments made by firms. But while that overall pattern was a fair description of reality about four decades ago, times have changed. Peter Chen, Loukas Karabarbounis and Brent Neiman describe the shift in patterns in \”The Global Rise of Corporate Saving,\” published as National Bureau of Economic Research Working Paper #23133 (February 2017). (NBER working papers are not freely available online, but many readers will have access via a library subscription.)

Here\’s the pattern of global saving by sector. Government saving relative to government GDP hovers just a bit above zero percent. But back in 1980, households saved about 14% of GDP while firms saved about 9%. Those shares have now flip-flopped, with firms now saving 13-14% of global GDP, and households saving 7-8% in recent years.

Meanwhile, investment across the sectors of the economy has remained much the same.

The authors sum up this change: \”In the last three decades, the sectoral composition of global saving has shifted. Whereas in the early 1980s most of investment spending at the global level was funded by saving supplied by the household sector, by the 2010s nearly two-thirds of investment spending at the global level was funded by saving supplied by the corporate sector. The shift in the supply of saving was not accompanied by changes in the composition of investment across sectors. Therefore, the corporate sector has now become a net lender of funds in the global economy.\”

This change is not wholly unexpected, in the sense that it fits with other economic patterns that have been noted, like high levels of corporate profits and high levels of household borrowing. Sorting out the reasons why corporate savings have become so large are still being investigated. But at this stage, the authors point out that the pattern is theoretically consistent with a model that includes lower levels of real interest rates, lower prices for investment goods, and lower corporate income tax rates.

This shift in the source of global saving represents a shake-up in the global financial system. For example, it means that firms become less likely to turn to external capital markets when raising money, because they can use their own savings instead.  At some ultimate level, of course, firms are owned by people, like shareholders and other owners. Thus, one way to describe this shift is that the underlying parameters of the global economic system have shifted so rather than households saving funds directly, households now use firms as the mechanism through which they save.

I\’m not opposed to spending more money on fixing up roads and bridges and other physical infrastructure–indeed, it\’s often an investment fully justified by cost-benefit analysis–but I am dubious that 21st century economic growth is going to be based on fewer potholes. When talking about investment to drive economic growth, I\’d like to see more focus on expansion of research and development spending.

The OECD recently updated its data on \”Main Science and Technology Indicators,\” and here\’s a figure generated from that website comparing R&D spending as a share of GDP in different places. The three countries for which there is data going back to 1981 are Germany (the purple line DEU), the United States (the olive-green line), and Japan (the red line). Notice that despite all the talk about how knowledge gains will be exceptionally important in the decades to come, US R&D spending as a share of GDP has barely budged in the last 35 years.

Of course, one can also point out that Germany\’s R&D as a share of GDP has barely budged either. And both the US and Germany have higher R&D spending, relative to GDP, than does the average for the 28 countries in the European Union (EU28, the yellow line).

On the other hand, R&D spending in Japan is higher than US levels. R&D spending in Korea (light blue line) has soared far above US levels. R&D spending in China (dark blue line) has risen to surpass EU28 levels and is moving closer to US and German levels.

Research and development efforts have spillovers that offer broader benefits across the economy. As I\’ve noted before, there are economic studies which suggest that optimal US R&D spending should be double or even quadrupled from its current levels. I\’d be happy to start with a smaller increase: say, taking steps to raise R&D spending by 1 percentage point of GDP in the next 5-10 years. For other posts on this issue, see \”US R&D in (Troubling) Context\” (February 25, 2015).

One of the gentle amusements from the hours I spend editing economics articles arises when authors forget the literal meaning of their own abbreviations. (Hey, there aren\’t a lot of belly-laughs in my job; you get your amusement where you can.) Here are five examples from early draft of papers.

An author writes about \”IT technology.\” Of course, IT means \”information technology,\” and no, the context was not about a highly specialized kind of second-order technology for information technology.

An author writes the \”CPI price index.\” Of course, CPI stands for Consumer Price Index.

An  author writes about \”the CAPM model.\” Of course, CAPM means \”capital asset pricing model.\”

An author writes about \”the VAT tax.\” Of course, VAT means \”value-added tax.\”

Au author writes about \”a major R&D research project.\” Of course R&D stands for \”research and development,\” and no, the author was not trying to describe a research project about R&D.

Acronyms and other specialized terminology always serve two functions. As economists and other specialists are quick to argue, they can be an exceptionally useful tool in professional communication, letting you refer to complex concepts in a compact form. But as economists and other specialists are slower to admit, the easy and casual use of acronyms and abbreviations is also a way of acting like part of an in-group, and of signalling to those outside the group your high-and-mighty status as an authority on the subject.

Again, I believe that both of these functions are always in play. One hopes that the first explanation based in the functionality gained from using acronyms and abbreviations will tend to predominate. One is always entitled to hope.

While the use of acronyms and abbreviations in economics may sometimes be overdone or even twee, at least little harm is done in the process (except for the pain suffered by hypersensitive editors). They are mostly just a sloppy signal that the author\’s brain isn\’t fully engaged and in a few cases can become comprehension-threatening, as the reader tries to figure out if the combination of abbreviation and repeated terms has some subtle meaning.

But misuse of acronyms and abbreviations can have bigger consequences. Text messaging has created a world where friendships and even true love can rise and fall based on the use of acronyms and emoticons. In medicine, for example, the National Coordinating Council for Medication Error Reporting and Prevention has an official list of abbreviations to avoid when writing prescriptions, because of the heightened risk of medication errors.  To make your blood run cold, here are a few of their examples:

Acronyms and abbreviations are powerful medicine, and should be used in limited doses–always remembering what you are using. It doesn\’t kill many more pixels to spell out terms.

China has 75% more capacity for generating wind power than does the United States, but it actuall produces 14% less wind power. Long Lam, Lee Branstetter, and Inês M. L. Azevedo explain the situation in \”Against the Wind: China’s Struggle to Integrate Wind Energy into Its National Grid,\” written as a \”Policy Brief\” for the Peterson Institute for International Economics (PB 17-5, January 2017). Here\’s a figure from their paper, showing China\’s capacity for wind generation outstripping the US, while the actual electricity generated falls short:

The authors write:

\”But close examination of China’s aggressive top-down approach to the promotion of renewable energy reveals that it has fallen far short of its ambitious goals. Turbines were quickly installed—but many of them were not connected to the power grid. After some turbines were connected, the state-owned enterprises (SOEs) that operate the national grid often refused to accept energy from them. These problems led to inefficiencies that are without precedent in the Western world. They help explain the shocking fact that although its installed wind energy capacity is 75 percent larger than that of the United States, China produces 14 percent less wind energy than the United States. Even in a political system with a strong centralized government, China’s push for renewable power faltered in the face of entrenched interests, weak incentives, and conflicting policy priorities. 

\”After accounting for the cost of building wind capacity that was not effectively utilized by the national grid, the cost of wind energy in China in the mid-2000s was twice as high as projected. A decade later costs had declined, but they were still 50 percent above projections. Consequently, the cost of carbon mitigation by replacing coal-generated electricity with wind energy has been four to six times higher than official estimates.\”

Apparently, up through about 2010, as much as 35% of China\’s wind capacity wasn\’t connected to the electrical grid at all. Since then, the electrical grid operators often \”curtail\” the electricity produced by wind power–which means they simply refuse to purchase it. The authors explain why:

\”Data from the first half of 2016 show a national curtailment rate of 21 percent—significantly higher than in any year between 2011 and 2015—and the highest curtailment rates typically lie in the second half of the year …  Curtailment rates are high and rising because keeping them high serves the financial interests of the grid companies. Since 2014 the growth of China’s energy-intensive industries has sharply decelerated, limiting electricity demand. At the same time, global coal prices have fallen sharply, lowering the cost of coal-powered electricity. Overestimating energy demand, the authorities permitted the construction of too many coal plants, forcing many of them to operate well below capacity. The intermittency of wind, and the need for the grid operators to purchase and dispatch the right amount of fossil energy–generated electricity to offset that intermittency, make wind power significantly more expensive for grid companies to use than coal power. Seeking to maximize their margins, the grid companies buy increasingly cheap coal energy and increase curtailment of wind energy.\”

China\’s wind power seems to be an example of how powerful government control can push in one direction, but even in this setting, economic realities will push back. I\’ve read a lot about China\’s efforts to expand renewable energy sources in the last 10-15 years, but  Lam, Branstetter, and Azevedo put the results to date in perspective when they write:

Back before the European Union became embroiled in how the euro was affecting trade balances and government borrowing, it was focused on a more basic project: reducing trade barriers between its members in the name of creating a single market. How\’s that going?

Vincent Aussilloux, Agnès Bénassy-Quéré, Clemens Fuest and Guntram Wolff offer an overview in \”Making the best of the European single market,\” written as a \”Policy Contribution\” for the Bruegel think tank (Issue No. 3, 2017). They argue that the gains from the European single market have been substantial, that productivity and investment are lagging in the EU, and that a renewed boost for the single market might be a big help.  I was particularly struck by their finding that trade across EU nations is at about one-fourth the level of trade across US states. They write (footnotes omitted):

\”Applying the synthetic counterfactuals method to various EU enlargements, Campos et al (2014) find that “per capita European incomes in the absence of the economic and political integration process would have been on average 12 per cent lower today, with substantial variations across countries, enlargements as well as over time”. This average figure is within the range found in the limited and fragile literature on this issue (5 to 20 percent, depending on the study). …

\”Still, trade between European countries is estimated to be about four times less than between US states once the influence of language and other factors like distance and population have been corrected for. For goods, non-tariff obstacles to trade are estimated to be around 45 percent of the value of trade on average, and for services, the order of magnitude is even higher. If the intensity of trade between member states could be doubled from a factor of 1/4 to a factor of 1/2 in order to narrow the gap with US states, it could translate into an average 14 percent higher income for Europeans (Aussilloux et al, 2011).\” 

The US has been experiencing a slowdown in productivity growth and in investment levels. In the EU, it\’s just as bad or worse. Here\’s a figure showing the productivity slowdown in the EU, with a few illustrative comparisons across countries. The productivity slowdown is everywhere, but it\’s worse in Europe.

One factor that is intertwined with Europe\’s productivity slowdown is its investment slowdown. Notice that for the EU as a whole, the levels of saving (blue line) and investment (red line) more-or-less track each other from 2000 up through 2011. But after 2011, saving rises and investment drops. In other words, there is capital being saved in the EU, but it\’s being invested somewhere else (as illustrated by the rise in the current account balance.)

Aussilloux, Bénassy-Quéré, Fuest and Wolff note that the problem for additional moves to a single market is that many of the easier steps have been taken. Thus, they propose that the next steps should focus on a relatively small number of key industries where economies of scale and trade might be especially productive. They write:

\”The extensive literature on how the single market could be deepened generally concludes that the easy gains have already been secured. The remaining barriers to trade are now in the services sectors and are much more difficult to eliminate, since services are and should be regulated: health care, legal services or data-intensive industries all need proper regulation. Since discrimination between nationals and non-nationals has already largely been eliminated, the challenge now is to harmonise regulations so that companies can develop their activities across borders in the same smooth way as they do within a country. …

\”Despite much talk and some relative successes – for example in the air transport sector – many of the most prominent services sectors remain fragmented. This is the case in the energy sector, rail transport, telecoms, consumer insurance markets, banking and professional services, among others. Although the big players in each of these sectors have activities in several EU countries, they operate not as if there was one single market, but on a series of distinct national markets.

\”The very slow progress in the pan-European integration of these sectors over the last 20 years suggests that a new approach is needed. For sectors with strong cross-border externalities and/or the potential for large economies of scale, the EU could define a single rule book and establish a single regulator or a network of national regulators, similarly to competition authorities. In networks, the national regulators would abide by the same rules, the same principles and methods, and by the same jurisprudence under the supervision and the coordination of a European regulator. This would be compatible with different national policies in certain areas, such as the choice of different energy mixes. Creating larger and more integrated markets is particularly important in the digital sector.\”

The authors also have discussions of various other possible steps, like a common business registration system across the EU countries, creating a \”common consolidated corporate income tax base\” to make it easier for companies to deal with varying tax laws across countries, and ways to better coordinate rules across countries about environmental protection, unemployment insurance, and social security. But the overall message is that despite a few decades of effort, and wave upon wave of rules that have often been about smalls-scale details, the EU remains a long way from a \”single market\” in big industries that matter.

For an earlier post on this topic, see \”What about the EU Single Market?\” (April 22, 2015).

We seem to be surrounded by wave upon wave of new information and communications technology. However, measured rates of productivity growth rates have been slow for more than a decade, with the slowdown starting well before the Great Recession and continuing after its end, both in the US and around the world, Bart van Ark seeks to explain this situation in \”The Productivity Paradox of the New Digital Economy,\” which appears in the Fall 2016 issue of International Productivity Monitor (pp. 3-18). In  a nutshell, his answer is that \”the New Digital Economy is still in its `installation phase\’ and productivity effects may occur only once the technology enters the `deployment phase\’.\”

At present, it\’s not just that productivity growth has slowed down, but counterintuitively, the sectors of the economy that make the biggest use of information and communications technology have been leading the way in this slowdown. Van Ark writes:

\”What\’s more, we find that when looking at the top half of industries which represent the most intensive users of digital technology (measured by their purchases of ICT [information and communications technology] assets and services relative to GDP) have collectively accounted for the largest part of the slowdown in productivity growth in all three economies since 2007, namely for 60 per cent of the productivity slowdown in the United States, 66 per cent of the slowdown in Germany, and 54 per cent of the slowdown in the United Kingdom. In the United States the contribution of the most intensive ICT-using industries declined from 46 per cent to 26 per cent of aggregate productivity growth between both periods. … The fact that ICT intensive users account for a larger part of the slowdown than less-intensive ICT users is another indication that the difficulty of absorbing the technology effectively is part of the explanation for the productivity slowdown.\”

Van Ark does believe that the growth of real output in information and communications technology is understated in the official statistics. But his broader theme is that information and communications technology is still in its \”installation stage,\” not its \”deployment stage.\” Here\’s how he sees the difference.

\”This article has argued that there are good reasons to believe that the New Digital Economy is still in the installation phase producing only random and localized gains in productivity in certain industries and geographies. … [W]we do not expect large aggregate growth effects from the New Digital Economy any time soon …\” 

As an example of where the installation phase is still taking place, van Ark points to digital services and \”big data\” projects:

\”[T]he shift toward full usage of digital services is incomplete as yet. A recent survey of more than 550 companies in Europe and the United States suggests only a modest uptake on one major usage of digital services, which is \”big data\” analytics. Only 28 per cent of companies in North America and 16 per cent in Europe had undertaken big data initiatives as part of their business processes in 2015. Another 25 per cent of companies in North America and 23 per cent in Europe had implemented a big data initiative as a pilot project. Hence about half of companies surveyed had not yet undertaken any big data initiative. Strikingly, the study also found that manufacturing companies were lagging in applying big data analytics projects in regular business processes by 14 percentage points relative to the retail sector (27 per cent versus 13 per cent of companies in each sector).\” 

An earlier post on \”When Technology Spreads Slowly\” (April 18, 2014) offered some discussion of transformative technologies that took decades to spread, with a focus on tractors and electrification.

I would be remiss not to mention that several other articles in this issue are worth particular attention, too, For example, Daniel Sichel reviews Robert J. Gordon\’s book, The Rise and Fall of American Growth,, and Gordon offers a response. Later in the same issue, Nicholas Oulton writes about \”The Mystery of TFP,\” which stands for total factor productivity: \”In all countries resources have been shifting away from industries with high TFP growth towards industries with low TFP growth. Nevertheless structural change has favoured TFP growth in most countries. Errors in measuring capital or in measuring the elasticity of output with respect to capital are unlikely to substantially reduce the role of TFP in explaining growth. The article concludes that the mystery of TFP is likely to remain as long as measurement error persists.\”